Anthropic S-1A warns it may soon meet its own RSI threshold

What happened

Anthropic's revised S-1A filing says the company may imminently meet its Responsible Scaling Policy definition of dramatically accelerated progress attributable to AI, though no current model has crossed that threshold. The filing also says Anthropic has in the past used AI to develop models whose capability leaps it did not recognize at the time, and that it is considering policy updates to keep humans in control of frontier model development.

Read next

• More changes in today’s briefingOpenAI’s updated Model Spec explained through its five core values

Reported by 2 sources

The story so far

Anthropic’s revised S-1A says its internal estimates and expectations for future models suggest it may soon meet its Responsible Scaling Policy threshold for “dramatically accelerated progress attributable to AI.” The company has not designated any current model as crossing that threshold and says it is still assessing the data. Anthropic also says its increasing use of AI in research and development could make model capabilities advance faster than it can develop, validate and implement safety, alignment and security measures. It says it does not expect to safely use Claude to design and build successive generations of frontier models fully autonomously without meaningful human supervision or understanding, even if that becomes technically possible.